Gas turbine engines, which are used to power aircraft, ships and other vehicles, and to provide shaft power in stationary installations, typically include a compressor section, a combustion chamber and a turbine section. Depending on the particular application, additional flow components including an intake diffuser, power turbine and a nozzle may also be incorporated.
During operation of the engine, air is ingested by the compressor section, which typically includes alternating rows of rotating and stationary blades. After the compressed air exits the compressor, it is typically decelerated for pressure recovery and distributed to one or more combustion chambers. There fuel is injected into the air and the air/fuel mixture is burned at approximately constant pressure. After combustion, the hot combustion gases are expanded through the turbine section, which typically includes an entry nozzle followed by one or more series of rotating and stationary blade rows for extraction of mechanical work from the expansion of the hot gases. The extracted work is used to rotate a shaft which, in turn, drives the rotating blades for the compressor.
In a basic aircraft propulsion application, the hot gases are then further expanded and discharged through a nozzle at a high velocity, providing substantial thrust to the engine. In addition to turning the compressor blades, the turbine section may include a power section to drive other devices, such as a fan as in a turbofan engine, a propeller as in a turboshaft engine or an electrical generator as in a stationary power system.
The optimal thermodynamic temperature for combustion in a gas turbine engine corresponds approximately to a stoichiometric fuel/air ratio which would result in combustion gas temperatures on the order of 4000.degree. Fahrenheit or higher. Conventional gas turbines, however, typically operate at a much lower turbine inlet temperature, ranging from about 2200 to 3000.degree. Fahrenheit, depending on the specific application of the engine. This limitation in operating temperature is due primarily to the lack of materials for use in the turbine section that are capable of withstanding the high temperatures of stoichiometric or near stoichiometric combustion in addition to the tremendous stresses resulting from the high rotational speeds of the turbine section and the harsh oxidizing environment imposed by the combustion products. As a result, conventional gas turbines often employ expensive high temperature superalloys in the turbine sections and yet still operate at relatively low power per unit of airflow and low fuel efficiencies when compared to the results that could be achieved at higher levels of fuel/air stoichiometry.
Similarly, the development of advanced turbine engines operating at higher pressure ratios results in air temperatures in the high pressure stages of the compressor that may exceed allowable limits for desirable structural materials. Consequently, such engine designs may require the use of active cooling or expensive superalloys in portions of the compressor section as well.
In an effort to overcome these limitations, substantial efforts have been directed primarily to cooling parts of the turbine section to a temperature substantially below that of the hot combustion gases. Most gas turbine engines, for example, divert a flow of "cool" bleed air from the compressor (bypassing the combustor) to hollow turbine blades. The turbine blades further include a series of pin holes along the blade surface to permit the "cool" bleed air from the blade interior to flow onto the blade surface thereby providing a thin film of cool air over the blade's external surface. This flow of bleed air then mixes with the expansion flow of hot combustion gases in the turbine section and is discharged through the nozzle of the engine. This type of blade cooling typically lowers the temperature of the turbine blades to approximately 1800.degree. Fahrenheit or less at the conventional operating temperatures noted above.
This approach, however, has many drawbacks. First, the drawing off of bleed air lowers the operating efficiency of the engine. Specifically, since the bleed air used to cool the turbine blades was not burned during the combustion process, it provides much less expansion work in the turbine section than the main combustion gas flow. Nonetheless, the turbine expended work compressing this air. Consequently, an optimum allowable quantity of bleed air is defined by balancing the reduction in the level of expansion work performed in the turbine section with the benefit of higher combustion gas temperature. In addition this approach is typically unable to lower blade temperatures to the point where expensive superalloy materials in the turbine section are no longer necessary.
Other design approaches considered or under development include various forms of forced liquid or gas flow through the turbine blades from an external source. Both closed and open flow paths within the turbine blades have been explored, including water circulation and steam throughflow, similar to cooling with compressor bleed air. These approaches, however, are not attractive for flight vehicles because of the added mass of the coolant. They also introduce added complexity to the engine by requiring coolant delivery from a stationary supply to the rotating components, making them do undesirable for stationary or surface vehicle systems as well.
Cooling by fuel circulation through the turbine blades prior to its combustion is also under development. In these designs, heat is absorbed by the fuel as a simple thermal mass, through controlled endothermic decomposition, or by reforming, e.g., of methane with steam. Nonetheless, all of these options require fluid delivery to and removal from the rotating components, as well as specialized fuels, catalysts or thermal conditions.
Another system used to cool turbine blades, which is depicted in FIG. 1 and disclosed in U.S. Pat. No. 5,299,418, the entire disclosure of which is incorporated herein by reference, includes the use of a cooling fluid F disposed within a sealed, internal cavity 34 in each turbine rotor 18. The rotor 18 includes a hub or disc section 28 proximate to the axis of rotation A--A of the engine and a blade section 30 extending outwardly from and supported by the hub 28. The hot combustion gases typically flow past the blade section 30. The rotor 18 further includes a wall 32 which defines the cavity 34. The cavity 34, moreover, is divided into a condensing section 36 located at the hub 28 and a vaporization section 38 located out at the blade 30.
The cooling fluid F cools that portion of the rotor wall 32 in the blade section 30 through evaporation. The physical properties of the cooling fluid F are such that it vaporizes at a temperature suitable for the wall 32 to operate with the hot combustion gases expanding past the rotor 18.
More specifically, rotation of the rotor 18 generates a centrifugal field which causes the cooling fluid F, in liquid form, to flow in the direction of arrows 40. The rotor 18 includes a series of radially spaced capture shelves 46 disposed along the inner surface of the wall 32 in the vaporization section 38. Each capture shelf 46 includes a well 50 facing the axis of rotation A--A. The cooling fluid F flows along the inner surface of the wall 32 from the condensing section 36 toward the vaporization section 38 where it enters the well 50 of the first capture shelf 46.
Ideally, as more cooling fluid F enters the first capture shelf 46, the level of cooling fluid F in that shelf 46 rises until the level of cooling fluid F reaches an inner lip 48 of the cooling shelf 46. The cooling fluid F then begins to cascade over the lip 48 of the first capture shelf 46 into the well 50 of the second capture shelf 46 and so on until each capture shelf 46 theoretically contains a stable pool of cooling fluid F.
Due to the heat flux from the hot combustion gases flowing past the rotor wall 32 a portion of cooling fluid F in each capture shelf 46 evaporates, thereby removing heat from that portion of the rotor wall 32 adjacent to the particular capture shelf 46. As more cooling fluid F evaporates, the localized vapor pressure in the vaporization section 38 increases relative to the vapor pressure in the condensing section 36. This differential in vapor pressure causes the evaporated cooling fluid in the vaporization section 38 to be pumped radially inward against the centrifugal field into the condensing section 36 as shown by arrows 42. The vapor flows to the cooler inner surfaces of the wall 32 in the condensing section 36 where it condenses back into cooling fluid F, thus maintaining a continuous, self-regulating recirculating flow of liquid coolant to the hot blade structure and of vapor to the condensing section 36. Heat rejected by evaporated cooling fluid as it reliquefies in the condensing section 36 may be removed by a number of conventional methods.
By providing a sufficient number of appropriately spaced capture shelves 46 and a sufficient volume of cooling fluid F, it was anticipated that this design would result in a rotor 18 capable of withstanding the higher temperatures associated with more efficient gas turbine operation. However, analytical studies, which were subsequently confirmed through experimentation, have identified limitations based on the internal geometry disclosed in that patent. First, the cooling fluid F traveling along the inner, planar surface of the wall 32 of the condensing section 36, as shown by arrows 40, is rapidly accelerated due to the high rotational speed of the turbine. Thus, by the time the cooling fluid F reaches the first capture shelf 46, the cooling fluid F may be moving at a substantial velocity. Indeed, the velocity of the fluid F is so great that the fluid F is actually turned by and expelled from the first capture shelf 46, rather than being captured therein as a stable pool. Consequently, there may be insufficient cooling fluid F available in the first capture shelf 46 to evaporatively cool the corresponding portion of the rotor wall 32.
In addition, since the cooling fluid F may fail to collect in the first capture shelf 46, there may be little or no cascading of cooling fluid F from one capture shelf 46 to the next. Instead, the majority of cooling fluid F, having been expelled from the first shelf 46 toward the center of the rotor cavity 34, may disperse in an uncontrolled manner and collect at the tip of the rotor 18. Moreover, any cooling fluid F that cascades from one capture shelf 46 to a subsequent capture shelf 46 also may be rapidly accelerated as it flows radially outward under the centrifugal field toward the subsequent capture shelf 46. As a result, this fluid F similarly may be expelled from the subsequent shelves 46 resulting in poor coolant distribution throughout the rotor 18.
The fluid F may also become separated from the rotor wall 32 thereby further disrupting the desired flow from one capture shelf 46 to the next. For example, a portion of fluid F flowing over the lip 48 of a capture shelf 46 may separate from the rotor wall due to the abrupt change in flow direction. Since it is the fluid's contact with the rotor wall 32 that imparts rotational acceleration to the fluid F, such separation also reduces the radial acceleration of the fluid F and results in an undefined direction of flow. In addition, since the rotors 18 are typically mounted within the turbine section at a high angle of attack relative to the plane of rotation, coriolis forces acting on the fluid F during rotation of the rotor 18 cause additional separation of the cooling fluid F from the rotor wall 32. In particular, coriolis forces, which act normal to both the axis of rotation and the direction of apparent motion, may force the fluid F circumferentially away from the rotor wall 32. Without a stable pool of cooling fluid F in each capture shelf 46, the system's ability to extract a sufficient quantity of heat from the rotor wall 32 or to provide uniform cooling over the blade section 30 is reduced.